Simplified ScreenCapper Pro

Overview

Simplified ScreenCapper Pro is a modular and automated pipeline for processing video files into high-quality frames and images. This updated version of the original Simplified ScreenCapper includes enhanced output quality, advanced features, and greater customization options. The pipeline follows a step-by-step process with separate scripts for each stage, ensuring flexibility and ease of extension.

---

Features

• High-Quality Output: Improved precision for frame extraction and face recognition.
• Modular Design: Independent scripts for each step allow easy replacement or extension.
• Automation-Friendly: Supports manual execution or optional daemon automation.
• Archiving and Cleanup: Automatically archives intermediate results and cleans up temporary files after processing.
• Customizable Quality Metrics: Use SSIM and edge detection thresholds to filter out duplicates and blurry images.

---

Workflow Pipeline

1. Frame Extraction (frame_export.py)

• Input: Video files from 1.cap_input/.
• Output: Extracted frames in 2.cap_output/.
• Task Indicator: task_running_Cap01.

2. Frame Transfer (transfer-cap-face.sh)

• Input: Frames from 2.cap_output/.
• Output: Frames moved to 3.face_input/.
• Archive: Zipped frames saved in archiv/1.cap/.
• Task Indicator: task_running_TransferCapToFace.

3. Face Recognition (face_recognition.py)

• Input: Frames from 3.face_input/.
• Output: Detected faces saved in 4.face_output/.
• Model: YOLOv8 (AniRef40000-l-epoch50.pt).
• Model Source: The YOLOv8 model must be downloaded from AniRef YOLOv8 GitHub Repository. Ensure you download the AniRef40000-l-epoch50.pt file and place it in the 00.scripts/model/ directory.
• Task Indicator: task_running_Face01.

4. Face Transfer (transfer-face-quali.sh)

• Input: Detected faces from 4.face_output/.
• Output: Faces moved to 5.quali_input/.
• Archive: Zipped faces saved in archiv/2.face/.
• Task Indicator: task_running_TransferFaceToQuali.

5. Quality Control (quality_check.py)

• Input: Faces from 5.quali_input/.
• Output: Filtered, high-quality images in 6.quali_output/.
• Task Indicator: task_running_QualityCheck.

6. Final Transfer and Cleanup (transfer-output.sh)

• Input: Final images from 6.quali_output/.
• Output: Images moved to output/.
• Archive: Zipped final images saved in archiv/3.quali/.
• Cleanup: Clears all process folders (1.cap_input to 6.quali_output).
• Task Indicator: task_running_TransferOutput.

7. Queue Pruning (pruning.sh)

• Purpose: Manually clear all queue folders.
• Safety: Asks for confirmation before deleting files.

---

Folder Structure

<Base Directory>/
├── 00.scripts/                # Contains all scripts
│   ├── master.sh              # Master script to control the pipeline
│   ├── frame_export.py        # Frame extraction
│   ├── face_recognition.py    # Face recognition
│   ├── quality_check.py       # Quality control
│   ├── transfer-cap-face.sh   # Transfer: CAP -> FACE
│   ├── transfer-face-quali.sh # Transfer: FACE -> QUALI
│   ├── transfer-output.sh     # Transfer: QUALI -> OUTPUT
│   ├── pruning.sh            # Clear all queue folders
│   └── model/                 # YOLOv8 model file
│       └── AniRef40000-l-epoch50.pt
├── 1.cap_input/               # Input folder for raw videos
├── 2.cap_output/              # Extracted frames
├── 3.face_input/              # Frames for face recognition
├── 4.face_output/             # Detected faces
├── 5.quali_input/             # Faces for quality control
├── 6.quali_output/            # High-quality output
├── output/                    # Final processed images
└── archiv/                    # Archive folders
    ├── 1.cap/
    ├── 2.face/
    └── 3.quali/

---

Installation

Prerequisites

• Python 3.8 or higher
• Required Python libraries:

  pip install opencv-python-headless tqdm ultralytics scikit-image

• System packages:

  sudo apt update
  sudo apt install zip unzip

Automated Installation with install.sh

To simplify the setup process, an install.sh script is provided. This script will:

1. Create the necessary folder structure.
2. Move all pipeline scripts into the 00.scripts directory.
3. Install required Python libraries.
4. Install necessary system packages.
5. Set executable permissions for the scripts.

Steps

1. Clone the repository:

   git clone https://github.com/AsaTyr2018/Simplified-ScreenCapper-Pro
   cd Simplified-ScreenCapper-Pro

2. Run the installation script:

   ./install.sh

3. Follow the prompts:

   • You can specify a custom base directory or use the current directory as the default.

4. Download the YOLOv8 model:

   • Visit the AniRef YOLOv8 GitHub Repository.
   • Download the AniRef40000-l-epoch50.pt model.
   • Place the model in the 00.scripts/model/ directory.

5. The installation script automatically sets execute permissions for all scripts.

---

Running the Pipeline

Manual Execution

Run the pipeline manually using:

<Base Directory>/00.scripts/master.sh

Optional Daemon

You can optionally set up a systemd daemon to automate the pipeline. The daemon will check the 1.cap_input/ folder every 5 minutes and trigger the pipeline if files are detected. Refer to the import-checker.timer and import-checker.service configuration for details.

---

Planned Features

• Classifier Integration: Add a character classifier for identifying specific characters within frames.
• Tagging System: Automate metadata tagging for processed images.

If you are experienced in these areas and want to contribute, feel free to open an issue or submit a pull request!

---

License

This project is licensed under the GNU General Public License v3.0.